To better understand and design plastic upcycling technologies, the development of toolkits to accurately determine transport and thermodynamic properties using only molecular structure and composition is essential. In this context, polyethylene constitutes the major fraction of global plastic waste and an excellent model material for the development of structure-composition-property relationships. In this work, we consider the two upcycling solutions: a) hydrocracking of polyolefins for chemical recycling and b) the design of circular polyethylene-like materials with built-in cleavable chemical linkages. For both applications, available databases for targeted properties are lacking, and quantitative property relationships are unknown. We integrate high-throughput molecular simulations and machine learning techniques into an automated pipeline that allow us to explore the parameter space of polyethylene-like structures and their blends to probe transport and thermodynamic properties in a consistent manner. The resulting data set enables the development of accurate predictive models for polyethylene-like materials as illustrated in the figure. Our results show how this workflow can be used to accelerate the design of sustainable solutions for plastic innovation.

Machine learning-assisted prediction of transport and thermodynamic properties of polyethylene-like materials

Fall 2022 abstracts, posters, and presentations will launch in mid-September.

ACS Fall 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.